Fire trucks cause Fires - Correlation vs Causation
This week we will be looking at correlation vs causation.
I can picture all you ’non-math’ people falling asleep or clicking ‘close’ already. But please read a little bit more.
We’ve all seen those articles in the media. Things like:
‘Successful people wake up at 4am…Why waking up early will make you more sucessful’.
‘Messy people earn less money…Tidy up an earn more’.
Other ones take the form of ‘Research finds X is linked to Y’ or ‘Scientists find a connection between X and Y’.[i]
Or more blatantly, ‘X causes Y’.
This is especially prevalent in so-called ‘health journalism’. If a study found a correlation between eating biscuits and cancer, headlines will often read something like, “Biscuits cause Cancer!”.[ii]
Let us use a more realistic example.
Eating breakfast has long been correlated with doing well at school. From this many journalists conclude that by feeding kids breakfast they will do better at school (breakfast causes academic success). However, while this could be true it is not necessarily true.[iii]
But what is the problem here?
Is the science wrong?
Maybe, but the main issue tends to be that journalists confuse correlation and causation.
What is the difference between correlation and causation?
One of my uni lecturers described this in a very clear way to me.
100% of house fires have fire trucks present therefore fire trucks cause house fires.
You can see the issue here. Just because there is a link between two things, doesn’t necessarily mean that one causes the other.
Correlation simply means that two things are related. If one goes up, the other goes up (positive correlation) or down (negative correlation). When this correlation is ‘statistically significant’ this means that the relationship between the two things is not likely due to pure chance.
However, as we have been harping on, correlation does not imply causation.
One possibility could be reverse causation. Let us look at the fire truck example. We know that it is far more likely that house fires actually cause the arrival of the fire trucks (through someone calling 000 – 911 for our US friends and 999 for our UK friends).
Another possibility could be a third underlying factor may cause both things. Let us look at the breakfast and school performance example from before. Perhaps the amount of money the family has may play a role. If a family has excess income, they may be able to afford breakfast and expensive schools and tutors. So income is the causal factor for breakfast and academic performance (Note: I have not done the research on this so this is purely for illustration purposes).
So what we need to establish causation then?
Well the bar is very high.
In psychology, three criteria are often stated (although there are other models with more criteria).
Firstly, we need to show correlation (that two are linked).
Secondly, we need to make sure there is a time relationship. One thing needs to come before the other.
Lastly, we need to rule out the fat they are caused by the third variable.[iv]
How do we establish causation?
The best way to achieve this is through something called a ‘Random Controlled Trial (RCT)’.
Photo by Olivier Verriest/iStock / Getty Images
You get a group of people, assign them to groups at random and not let them or the researchers know who is who (this is called double blinding).
You give one group an intervention and the other a placebo so they don’t know who got the intervention and who didn’t.
You the analyse the results to see if there was a change and a difference between the groups.
Of course, this is easier said than done and can be unethical.
Say we wanted to use an RCT to determine if smoking caused cancer. We would have to get one group to smoke (causing s blinding issue and an ethical issue) and see if more of them got cancer (a huge ethical issue).[v]
So observational studies are conducted where group of people are observed over time to try and establish causality, but it is still extremely difficult to do in may causes.[vi]
What does this mean to me?
This doesn’t mean the science is rubbish.
But it is important to ask yourself those question when you read an article that blatantly claims that one thing causes or another or you should change something to achieve something else.
Think about, are these things directly related. Which one causes the other? Could there be a third thing that causes them both?
Rather than just automatically believing, you and now do a bit of your own critical analysis and be better informed.
[i] https://www.bbc.com/news/magazine-27537142
[ii] https://www.theguardian.com/science/blog/2014/dec/10/science-health-news-hype-press-releases-universities
[iii] https://senseaboutscienceusa.org/causation-vs-correlation
[iv] https://www.sagepub.com/sites/default/files/upm-binaries/23639_Chapter_5_Causation_and_Experimental_Design.pdf
[v] https://www.abs.gov.au/websitedbs/a3121120.nsf/home/statistical+language+-+correlation+and+causation
[vi] https://senseaboutscienceusa.org/causation-vs-correlation/
Photo by umesh chandra/iStock / Getty Images
With things like ‘Zip’ and ‘AfterPay’ we can buy what we want and pay it off later.
Our culture encourages us to seek temporary solutions that we can do now such as ‘get-rich-quick’ schemes, fast fat loss programs and even the arguably over-prescription of drugs to quickly solve our issues.[i]
Photo by kentarus/iStock / Getty Images
For many of us in the Western world, we are so lucky to have these conveniences at our fingertips.
But what about delayed gratification, then?
What is Delayed Gratification?
It is as simple as it seems, it is the tendency to give up an immediate reward in order to pursue a long-term reward.[ii]
The Marshmallow Experiment
We cannot talk about delayed gratification without talking about the hallmark Marshmallow Experiment.
For those who haven’t heard of this, have a look at YouTube videos on this. They are a great watch.
Essentially, psychologist Dr Walter Mischel, PhD, played what may seem like a cruel experiment on some four year olds (keep reading, the kids aren’t hurt). Experimenters placed a marshmallow in front of them. They were told they could eat the one marshmallow now or wait until the experimenter returned. If they waited they would get TWO marshmallows.[iii]
The study then followed these kids to see if there were any differences between those who at the marshmallow and those who waited (showed delayed gratification).
The Benefits of Delayed Gratification
These follow-ups showed kids that waited for the second marshmallow were far more successful than those who ate the first marshmallow. They had:
Better social skills,
Responded to stress better,
Were less likely to be obese,
Had higher SAT scores, and
Even showed a more developed prefrontal cortex.[iv]
Other studies have found similar results with a study of high school teachers finding those who showed higher levels of delayed gratification had more job satisfaction and less stress.[v]
Can We Improve Our Ability to Delay Gratification?
While originally, the ability was seen as innate and a personality trait (we were born with it and it could not be changed). There has been some thought, more recently, that this might not be the case. Recent studies have shown that self-control exercises can lead to improvements in self-control.[vi]
These findings have shown that self-control makes it harder to exhibit self-control later (it tires like a muscle) and it is proposed that self-control exercise may improve the stamina of this ‘muscle’.[vii] There is still much to learn here, though.
Conclusion
Delayed gratification seems to be linked with many benefits and research is starting to suggest that we can actually improve out ability to do it.
How well do you exhibit self-control and delayed gratification? Would you eat the marshmallow straight away or wait to get two?
[i] https://www.psychologytoday.com/au/blog/your-emotional-meter/201712/the-benefits-delaying-gratification
[ii] Gerhart, J. I., Heath, N. M., Fitzgerald, C., & Hoerger, M. (2013). Direct and indirect associations between experiential avoidance and reduced delay of gratification. Journal of Contextual Behavioral Science, 2(1-2), 9-14.
[iii] https://www.apa.org/monitor/2014/12/marshmallow-test
[iv] https://www.psychologytoday.com/au/blog/the-age-overindulgence
[v] Mohsin, F. Z., & Ayub, N. (2014). The relationship between procrastination, delay of gratification, and job satisfaction among high school teachers. Japanese Psychological Research, 56(3), 224-234.
[vi] Muraven, M., Baumeister, R. F., & Tice, D. M. (1999). Longitudinal improvement of self-regulation through practice: Building self-control strength through repeated exercise. The Journal of social psychology, 139(4), 446-457.
[vii] Ibid.